- Mar 06, 2010
-
-
Mark Brown authored
I2C only at the minute. Signed-off-by:
Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by:
Liam Girdwood <lrg@slimlogic.co.uk>
-
- Mar 04, 2010
-
-
Mark Brown authored
Report the current FIFO depth when delay is queried. The FIFO is only 16 frames deep so the latency will be at most a couple of miliseconds (and we tend to end up reporting zero most of the time) but it may help some applications. Signed-off-by:
-
Mark Brown authored
The WM8960 headphone outputs can be run in capless mode with OUT3 used to drive a pseudo ground for the headphone drivers. In this mode the mono mixer is not used, the mixer should be turned on in concert with the headphone output drivers and the device bias levels are managed differently. Also tweak the existing bias management to remove the use of active discharge while we're at it since that's often audible. Signed-off-by:
-
Mark Brown authored
Avoids machine files having to peer into sound/soc which is a bit rude and icky. Signed-off-by:
-
Mark Brown authored
The driver name gets used by dev_() logging so use something a bit more idiomatic. Signed-off-by:
-
Peter Ujfalusi authored
The delay callback can be used by the core to query the delay on the dai caused by FIFO or delay in the platform side. In case if both CPU and CODEC dai has FIFO the delay reported by each will be added to form the full delay on the chain. If none of the dai has FIFO, than the delay will be kept as zero. Signed-off-by:
Peter Ujfalusi <peter.ujfalusi@nokia.com> Acked-by:
-
Peter Ujfalusi authored
Create a soc level wrapper for pcm_pointer callback. This will facilitate the soc level handling of different HW buffers in the audio path. Signed-off-by:
-
Peter Ujfalusi authored
My editor removes the tailing spaces, which causes problems when changing the soc-core.c Removing the space. Signed-off-by:
-
Mark Brown authored
Initial support for audio using the 1133-EV1 audio and PMIC module for the i.MX31ADS. Currently only playback is supported, and the FIQ DMA driver has performance problems at higher sample rates which cause audible dropouts. This driver is based heavily on an out of tree one written by Liam Girdwood. Signed-off-by:
-
- Feb 26, 2010
-
-
Jassi Brar authored
If we are to have a snd_soc_dai i.e, cpu_dai and codec_dai, shared among two or more dai_links we need to log the number of active users of the dai. For that, we change semantics of the snd_soc_dai.active flag from indicator to reference counter. Signed-off-by:
Jassi Brar <jassi.brar@samsung.com> Acked-by:
-
- Feb 22, 2010
-
-
jassi brar authored
In order for having snd_soc_dais shared among two or more dai_links, remove the relatively global runtime field from the struct snd_soc_dai Signed-off-by:
-
jassi brar authored
Currently only the atmel driver make use of snd_soc_dai.runtime field. If the dais are to be shared among two or more dai_links, the field must be got rid of. So, in atmel driver reach the substream via dai_link->pcm so as to not depend of snd_soc_dai.runtime field. Signed-off-by:
-
jassi brar authored
Passing pointer to relevant dai_link provides easier reach to the ASoC tree in suspend/resume of snd_soc_platform. It also provides direct access to the dai at the other end of the dai_link. Signed-off-by:
-
- Feb 17, 2010
-
-
Mark Brown authored
Fixes a warning. Signed-off-by:
-
Peter Ujfalusi authored
Disable the amplifiers for the headset outputs, and do not select routings by default to the headset outputs. Signed-off-by:
-
Peter Ujfalusi authored
OSCSET calculation was not correct in case of 44.1KHz sampling rate. With small adjustment both 48 and 44.1 KHz calculation now gives the correct value. Signed-off-by:
-
Peter Ujfalusi authored
In repeated playback the FIFOFLUSH bit remained set, and never has been cleared. Clear it during the setup phase. Signed-off-by:
-
Mark Brown authored
Provide a sysfs file allowing userspace to inspect and change the pmdown_time setting at runtime. Signed-off-by:
-
Mark Brown authored
Make the pmdown_time a per-card setting rather than a global one, initialised before the card initialisation runs. This allows cards to override the default setting if it makes sense to do so (for example, due to an unavoidable pop). Signed-off-by:
-
- Feb 12, 2010
-
-
Mark Brown authored
The WM2000 is a low power, high quality handset receiver speaker driver with Wolfson myZone™ Ambient Noise Cancellation (ANC). It provides enhanced voice communication quality in a noisy environment if the handset acoustics are designed appropriately. Signed-off-by:
-
Guennadi Liakhovetski authored
Audio on Migo-R cannot work if CONFIG_SH_DMA_API=y, but compilation should not break anyway. Signed-off-by:
Guennadi Liakhovetski <g.liakhovetski@gmx.de> Acked-by:
-
Thomas Weber authored
This patch expands the omap3beagle sound soc for the beagle board clone DevKit8000. Signed-off-by:
Thomas Weber <weber@corscience.de> Acked-by:
Jarkko Nikula <jhnikula@gmail.com> Acked-by:
-
- Feb 10, 2010
-
-
Paul Menzel authored
Signed-off-by:
Paul Menzel <paulepanter@users.sourceforge.net> Acked-by:
-
Peter Ujfalusi authored
The serial interface (TDM/I2S) for the audio block have been constantly enabled. Introduce a new DAPM_SUPPLY for handling the AIF_EN bit, so the interface is only enabled, when there is a need for it. For example when only the analog loopback is enabled, there is no need to keep the serial interface active. I have added the persons who contributed to the Voice path of twl4030 codec driver, so they might have the ability to test this patch, and send an update for the Voice path, if it is necessary Signed-off-by:
-
Daniel Mack authored
Enable the bulk regulators at probe time so we can safely disable them again when going to suspend without confusing the reference counter. Signed-off-by:
Daniel Mack <daniel@caiaq.de> Acked-by:
-
Mark Brown authored
This patch adds sound support for Phytec PhyCORE / PhyCARD modules in AC97 mode. Signed-off-by:
Sascha Hauer <s.hauer@pengutronix.de> Acked-by:
-
- Feb 06, 2010
-
-
Grazvydas Ignotas authored
Pandora's external DAC is connected to VSIM TWL4030 supply, so let's start switching it too to save more power. Also DAC got it's own DAPM handler. Signed-off-by:
Grazvydas Ignotas <notasas@gmail.com> Acked-by:
-
- Feb 05, 2010
-
-
Mark Brown authored
-
Grazvydas Ignotas authored
Pandora's external DAC is using 256*Fs output from the TWL4030 codec, and TWL4030 needs to have APLL enabled for it's 256*Fs output to function. Signed-off-by:
-
- Feb 04, 2010
-
-
Peter Ujfalusi authored
The module unloading path had several problems: - it freed up the private structure twice - it freed up the codec structure, which was allocated as part of the private structure - it did not freed up the reg_cache - it did not unregistered the dais and the codec Signed-off-by:
-
Mark Brown authored
The WM8912 is a DAC only device register compatible with the WM8904 CODEC with ADC portions omitted. Support it within the WM8904 driver based on the configured I2C device name. Signed-off-by:
-
Mark Brown authored
Handle the output PGAs as part of the output powerup since they can never be powered separately and reorder things so that we remove the output shorts after both line and headphone outputs have been brought up, minimising the opportunity for any issues. Signed-off-by:
-
Mark Brown authored
As well as disabling the biases of the CODEC the drop into BIAS_OFF will also disable all the regulators powering the CODEC, allowing even greater power savings on appropriately configured systems. Since the regulator API does not currently provide notification when regulators are disabled we assume that this always happens when we stop using the regulators. Once 2.6.34 is merged this code can be optimised to only sync the cache when power was actually removed. Signed-off-by:
-
Mark Brown authored
While the regulators are disabled we cache all register writes. Currently we assume that the regulator disable actually takes effect, after the merge with the regulator tree in 2.6.34 the regulator API will be able to notify us if the power is actually removed (due to constraints or regulator sharing it may not be). Signed-off-by:
-
Mark Brown authored
At the minute the regulators are simply enabled for the entire lifetime of the device. Signed-off-by:
-
Mark Brown authored
Saves a little bit of code duplication. Signed-off-by:
-
Mark Brown authored
Add a bit to the CODEC structure indicating if a cache sync is required. By default this will be set if a cache only write is done to a soc-cache register cache. This allows us to avoid syncing the cache back after using cache only writes if there were no changes. Signed-off-by:
-
Mark Brown authored
Currently the soc-cache code will always write to the device, meaning that we need the device to be powered and active at pretty much all times the system is active. Allowing cache only writes lays some groundwork for future enhancements to allow devices to be put into a full off state when the audio subsystem is idle. Signed-off-by:
-
Guennadi Liakhovetski authored
ctrl_outl() has become void at some point, which breaks compilation of fsi.c. Make writing functions void, as their output is anyway not evaluated, and use __raw_writel and __raw_readl instead of deprecated ctrl_outl and ctrl_inl respectively. Signed-off-by:
-
- Feb 02, 2010
-
-
Daniel Mack authored
Unconditionally save the register states when suspending and restore them again at resume time. Register contents were not preserved over suspend, and hence the driver takes false assumptions about them. The clock must be enabled to access the register block. Signed-off-by:
-
